利用自動對焦重建微波影像

Abstract

微波影像在遙測地表、獲得地面上土壤、水資源等，甚至在軍事方面都是相當重要的影像資訊。由於微波影像的同調性質，需要相當精確之相位計算，稍許的相角誤差會造成影像模糊、解析度降低，而使其價值降低。這裡所提及的相位誤差皆指方位角方向的誤差，並且是因為飛機路徑偏移所造成的。自動聚焦演算法是一種改善微波影像品質的演算法，可以估測訊號的相位誤差，以達到消除相角誤差，改善影像之解析度。 自動聚焦技術主要包含兩個部分：誤差估測和誤差補償。傳統由感測器執行之運動補償功能只能測量到相位的一次項係數，如果加上自動聚焦法，則可以將二階以上的誤差項消除。本篇論文主要是探討在不同的模式下，自動聚焦演算法的成效。利用自動聚焦演算法來補助系統的運動補償功能或成為微波影像系統之一部分架構，以期改善微波影像之品質。

Microwave images are important for acquiring the surface data of the earth such as the soil of the ground, water resource, and the applications in military. Because of the coherent characteristics of microwave, the phase must be calculated accurately. A little phase error will cause image degrading and useless. Phase errors mentioned here mean errors in azimuth direction. Autofocus technique is an algorithm to improve microwave images quality. Autofocus technique can estimate phase errors and remove phase errors to improve the images resolution. Autofocus technique generally encompasses two steps：phase estimation and phase compensation. First order coefficient of the phase error function can be estimated by sensor-based motion compensation. Second or higher order term of phase error function could be estimated and cancelled by autofocus technique. In this paper, the effect of the autofocus algorithm in different modes is presented. The autofocus algorithm is used to assist the capability of motion compensation of the system. And the autofocus algorithm is expected to improve the quality of microwave images.